Since the introduction of recent regulations, the expectations for Apple’s compliance in China have become increasingly stringent. Algorithms influencing public visibility online or through artificial intelligence are required to be registered with Chinese authorities. New AI legislation primarily targets controlling public-facing models that Western technology companies are keen to engage with.
Experts, such as Tan, note the necessity for companies to file with regulators, possibly involving the disclosure of detailed information like coding, which many tech companies might be reluctant to provide. The enforcement of such measures is facilitated by the prevailing power dynamics, which increasingly favor China. According to Tan, China is no longer merely following in technological advancements but is leading in several areas.
From a Western perspective, China’s regulations for generative AI present a mix of admiration and concern. Vague censorship requirements within the regulation, such as ensuring content aligns with the correct political direction and does not disrupt economic or social order, are highlighted in a Carnegie Endowment paper. “Deep synthesis,” the term used by China’s Cyberspace Administration (CAC) for generative AI, faces restrictions preventing references to the Dalai Lama, Taiwan as a separate nation, or the Uyghurs.
Western language models (LLMs) are generally less constrained, making the adaptation to China’s rigorous standards challenging. However, Chinese tech firms have successfully adhered to these restrictions, as evidenced by the approval of 188 LLMs for use by August 2024, a significant increase from January 2024, reported by the South China Morning Post.
Incorporating a customized version of an LLM to comply with Chinese regulations could be seen as routine for Apple. The company already censors its app store to align with China’s policies and collaborates with local entities. However, with Apple Intelligence’s generative AI becoming integral to iPhones and other devices, concerns arise about Apple’s potential alignment with Chinese state interests, which could be unsettling for a U.S. company.
In August, Zhuang Rongwen, head of the Cyberspace Administration of China, stated that generative AI, such as chatbots, is significantly driving economic and social growth. A 2021 report by The New York Times indicated that the Chinese government may not need data from Chinese iPhone users to surveil its citizens, having more robust methods in place. However, with the advent of generative AI, Apple might inadvertently play a more active role in supporting the Chinese Communist Party’s objectives.
